Amazon announced Amazon Bedrock AgentCore payments, a preview service that allows AI agents to transact on behalf of end users while implementing built-in guardrails to manage risks. The service is designed to address the challenges of autonomous payments, such as runaway spending and unauthorized access to financial information. Amazon Bedrock AgentCore payments is available in preview in US East (N. Virginia), US West (Oregon), Europe (Frankfurt), and Asia Pacific (Sydney). Features and APIs may change before general availability. The service integrates with Amazon Bedrock AgentCore to provide infrastructure-layer controls that ensure safety and compliance. It enables developers to define spending limits and session timeouts, ensuring that agents operate within set boundaries. The system also allows end users to retain control over their funds by requiring explicit delegation before any transactions can occur. 'Putting real money behind an autonomous system raises a new set of risks,' said Amazon. 'These include agents acting autonomously over long sessions, model non-determinism, and a wider exposure surface between agent code and the end user’s funds.' *Source: [awsml](https://aws.amazon.com/blogs/machine-learning/enable-safe-agentic-payments-with-built-in-guardrails-using-amazon-bedrock-agentcore-payments/)*